Investigation of the terminal P4 domain in a series of d-phenylglycinamide-based factor Xa inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.09.105

Several P4 domain derivatives of the general d-phenylglycinamide-based scaffold (2) were synthesized and evaluated for their ability to bind to the serine protease factor Xa. Some of the more potent compounds were evaluated for their anticoagulant effects

Full text of DOI:10.1016/j.bmcl.2007.09.105

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6910–6913
Investigation of the terminal P4 domain in a series
of D-phenylglycinamide-based factor Xa inhibitors
Jeffry B. Franciskovich,a,* John J. Masters,a Wayne W. Weber,a Valentine J. Klimkowski,a
Michael Chouinard,a Philip R. Sipes,a Lea M. Johnson,a David W. Snyder,a
Marcia K. Chastain,a Trelia J. Craft,a Richard D. Towner,a Donetta S. Gifford-Moore,a
Larry L. Froelich,a Jeffrey K. Smallwood,a Ronald S. Foster,a Gerald F. Smith,a
John W. Liebeschuetz,b,c Christopher W. Murrayb,d and Stephen C. Youngb,e
aLilly Research Laboratories, Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN 46285, USA
bTularik Ltd, Amgen, Aleutian House, Tytherington Business Park, Macclesfield, Cheshire, UK
cThe Cambridge Crystallographic Data Center, 12 Union Road, Cambridge CB2 1EZ, UK
dAstex Therapeutics, 436 Cambridge Science Park, Milton Road, Cambridge CB4 0WE, UK
eVertex Pharmaceuticals (Europe) Ltd, 88 Milton Park, Abingdon Oxfordshire OX14 4RY, UK
Received 12 July 2007; revised 30 August 2007; accepted 4 September 2007
Available online 24 October 2007
Abstract—Several P4 domain derivatives of the general D-phenylglycinamide-based scaffold (2) were synthesized and evaluated for
their ability to bind to the serine protease factor Xa. Some of the more potent compounds were evaluated for their anticoagulant
effects in vitro. A select subset containing various P1 indole constructs was further evaluated for their pharmacokinetic properties
after oral administration to rats.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Factor Xa (fXa) is a serine protease that plays a crucial
role in the blood coagulation cascade.1 FXa occupies a
common point in both the intrinsic and extrinsic path-
ways and is responsible for the production of thrombin.
Thrombin, another serine protease, converts fibrinogen
to fibrin, and fibrin cross-links with platelets eventually
leading to blood clot formation. Because fXa is essential
to this pathway, it has been hypothesized that the design
of a small molecule inhibitor of factor Xa would be use-
ful in treating patients with thromboembolic diseases.2
reasonable anticoagulant activity as measured by dou-
bling prothrombin time in vitro (2· PT).5
N
O
N
N
O
N
1
Kassa = 450 x 106 L/mol
2x PTb = 1.5 μM
a
Kass represents the apparent association constant, as measured by the
methods of Smith5 et al.
Previously, our laboratories reported the discovery of
D-phenylglycine-based inhibitors of fXa (compound
1).3 SAR studies carried out in the P1 portion of these
inhibitors led to the discovery that fused bicyclic ele-
ments enhanced binding affinity in vitro.4 Compound
1, which contains a 6-indolyl ring system, is a potent
inhibitor of factor Xa (Kass). In addition to having po-
tent binding affinity for fXa, compound 1 also displays
b 2x PT is defined as the concentration of compound required to double
the time to clot formation in the prothrombin time assay.
A crystal structure of 1 (Fig. 1) indicates the 6-indolyl
ring binds in the S1 pocket and the N-methylbispiperi-
dine domain binds in the S4 region of this protease.
R
O
N
N
O
Keywords:
Antithrombotics.
Factor
Xa
inhibitors;
D-Phenylgylcinamides;
N
2
*
Corresponding author. E-mail: j.franciskovich@lilly.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.09.105

J. B. Franciskovich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6910–6913
6911
The factor Xa binding affinity and anticoagulant po-
tency for compounds 3a–3l are listed in Table 1. Com-
pounds 3b and 3c with a tertiary amine have a higher
binding affinity than the secondary amine 3a. The pyr-
rolidine derivatives that contain a hydrophilic hydroxyl
group (3d–3g) are less active than the unsubstituted pyr-
rolidine 3c. Similarly, the piperidine derivatives contain-
ing a hydroxyl group 3i and 3j have lower binding
affinity than the unsubstituted piperidine 3h. The bind-
ing affinity for fXa is increased 2-fold when the ring size
is expanded from 5 to 6 atoms (3c vs 3h). Incorporation
of a second nitrogen atom into the piperidine ring re-
Table 1. Human factor Xa binding affinity and anticoagulant activity
of select inhibitors 3a–3l
R
O
N
N
O
Figure 1. Factor Xa complex with compound 1. X-ray crystal
˚
structure (1.90 A resolution).
N
a
Compound
R
Kass (·106 L/mol) 2· PTb (lM)
1
—
450
1.5
In this paper, we present the synthesis and terminal P4
domain SAR for a select set of compounds of formula
2 in which hydrogen bonding motifs are incorporated.
The effect on anticoagulant activity in vitro of com-
pounds displaying the highest binding affinity is pre-
sented. Additionally, a select subset of derivatives
containing specific P1 fused bicyclic constructs and their
binding affinity to fXa, anticoagulant potency, and oral
exposure properties are reported.
3a
3b
3c
3d
7.8
—
N
N
28
24
18
2.3
2.2
2.2
N
N
OH
The synthetic approach utilized for compounds 3a–3l is
shown in Scheme 1. The commercially available CBZ
protected D-phenylglycine is coupled with 4-piperidine-
methanol to afford 4 in high yield with minimal epimer-
ization. Removal of the CBZ group proceeds smoothly
to provide the free amine 5, and coupling with indole-
6-carboxylic acid affords 6 in modest yield. Mesylation
of 6 followed by displacement with a variety of amines
results in the final compounds.
3e
3f
9.7
—
—
N
OH
OH
5.0
4.0
N
(rac)
OH
3g
3h
3i
—
1.2
—
N
48
OH
N
OH
a
OH
N
+
OH
CBZN
CBZN
N
O
10
O
4
N
OH
O
OH
3j
0.7
—
OH
c
b
N
N
N
H2N
O
N
O
N
H
N
5
6
3k
3l
3.9
1.5
—
—
N
1
2
O
N
R R
N
d, e
N
N
N
O
N
Scheme 1. Reagents and conditions: (a) EDCI, HOAt, DCM/DMF
(1:1), 94%, >95% ee; (b) 10% Pd/C, H2, EtOAc/EtOH (1:1), 94%; (c)
DCC, HOBt, DMF, indole-6-carboxylic acid, 54%; (d) DCM, MsCl,
NEt3, 83%; (e) THF, NaI, K2CO3, NR1R2, 60 ꢁC, 25–55% yield.
a Kass represents the apparent association constant, as measured by the
methods of Smith et al.5
b 2· PT is defined as the concentration of compound required to
double the time for clot formation in the prothrombin time assay.

6912
J. B. Franciskovich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6910–6913
Table 2. Human factor Xa binding affinity, anticoagulant activity,
clogP, and oral exposure properties of various P1 indole constructs
compared to 1
sults in decreased affinity for fXa as shown in com-
pounds 3k and 3l.
A few compounds were evaluated for their anticoagu-
lant potency in human plasma. The compounds with
PT data show that as the binding affinity for fXa in-
creases, the concentration of compound required to
double the time for clot formation in plasma decreases.
Compounds 3c and 3h are representative of this trend.
In contrast, although 3h displays an order of magnitude
weaker binding affinity for fXa than 1, a slightly lower
concentration is required to double the clot formation
time in plasma.
N
O
N
N
O
X
N
Compound
X
H
Kass clogP 2· PT Cmax AUC (0–7 h)
(lM)a (lg h/ml)a
(lM)
3h
7
48 4.3
1.3
1.5
1.1
1.5
4.7
2.4
1.7
4.0
23
14
10
19
Me 110 4.8
8
Cl
—
710 5.2
450 4.2
1
The crystal structure of compound 3h is shown in
Figure 2. The extended P4 domain is shown to bind in
the S4 pocket as predicted. A possible explanation for
the decrease in activity for 3h may be that overall the
P4 group is positioned in a less favorable arrangement
than the equivalent group in compound 1 (Fig. 1).
a Cmax and AUC values were pharmacodynamically determined mea-
suring anti-Xa equivalents in rat plasma. Animals were dosed by oral
administration at 20 mpk (n = 3).
affinity of 2.6-fold and 17-fold, respectively. However,
the anticoagulant potencies of 7 and 8 are similar to that
of compound 3h. This reduced anticoagulant activity
could be due to the increased lipophilicity of these mol-
ecules as shown by the calculated clogP values. The oral
exposure of 3h is greater than those for compounds 7
and 8. In addition, the plasma exposure of compound
3h is greater than that of 1 and has a similar clogP
value.
Specifically, in 3h the lower piperidine ring is positioned
perpendicular with respect to the face of the Trp215 in-
dole sidechain, instead of laying directly on it as in 1.
This will reduce the hydrophobic contact between the li-
gand and the aromatic residues in this region. Addition-
ally, the potentially charged amino group appears to be
partially buried and less solvent exposed in 3h compared
to 1.
In summary, a series of D-phenylglycine-based inhibitors
modifying the terminal P4 element were synthesized and
identified as fXa inhibitors. The SAR indicates that
extension of the P4 element with or without rings con-
taining hydrogen bonding elements does not increase
factor Xa binding affinity relative to 1. However, the
study results in the identification of 3h, a compound that
displays similar anticoagulant activity to 1 while pos-
sessing an order of magnitude less binding affinity. In-
creased affinity for fXa is achieved by incorporating a
3-substituent on the indole P1 domain of this series.
The oral exposure of 3h is found to be greater than those
of 7 and 8, but similar to 1. Within this series, com-
pounds that display lower exposure tend to be more
lipophilic as determined by clogP values.
We combined the most potent P4 element in this SAR
with a couple of 3-substituted indole P1 constructs.
Preparation of compounds 7 and 8 was carried out
utilizing the same chemistry as in Scheme 1. The binding
affinity, anticoagulant potency, clogP,6 and oral expo-
sure data7 compared to 1 and 3h are shown in Table 2.
Addition of a methyl group or a chlorine atom in the 3
position of the indole results in increased fXa binding
Acknowledgments
The author thanks Joe Klimkowski, John Masters, and
Thomas E. Jackson for helpful discussions and review of
this manuscript.
References and notes
1. Coleman, R. W.; Marder, V. J.; Salzman, E. W.; Hirsch, J.
In Hemostasis and Thrombosis: Basic Principles and Clinical
Practice; Coleman, R. W., Marder, V. J., Salzman, E. W.,
Hirsh, J., Eds., 3rd ed.; J.P. Lippincott: Philadelphia, 1994;
pp 3–18.
2. (a) Quan, M.; Smallheer, J. Curr. Opin. Drug Discov. Dev.
2004, 7, 460; (b) Walenga, J. M.; Jeske, W. P. Curr. Opin.
Invest. Drugs 2003, 4, 272; (c) Gould, W. R.; Leadley, R. J.
Curr. Pharm. Design 2003, 9, 2337; (d) Walenga, J. M.;
Figure 2. Factor Xa complex with compound 3h. X-ray crystal
˚
structure (2.40 A resolution).

J. B. Franciskovich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6910–6913
6913
Jeske, W. P.; Hoppenstedt, D.; Fareed, J. Meth. Mol. Med.
2004, 93, 95.
purity, and dissolved in GC–MS or HPLC grade meth-
anol (OD at 350 nm < .001) with brief sonication in a
water bath. Compounds were serially diluted in methanol
and aliquoted into pooled plasma from vehicle treated
animals in triplicate, for use as calibration standards. A
solution for human factor Xa was prepared by dissolving
18.8 mg of fatty acid-free bovine serum albumin (Sigma
RIA grade A-7888) in 15 ml of 1· TBS. A solution of Xa
substrate, S-2222 (MidWest Bio-Tech, South Bend, IN),
was diluted to 0.5 mg/ml in 18.2 MX purified water.
Water (5 ll) was added to all wells reserved for
standards. Methanol (5 ll of 50%) was added to wells
reserved for samples of unknown compound level. Ex
vivo plasma (25 ll) from animals dosed with compounds
was thawed at 4 ꢁC. Samples and calibration standards
were mixed, incubated in a 37 ꢁC water bath for 15 min,
and diluted into individual wells of a 96-well, polystyrene,
uncoated assay plate containing 50 ll of 3· TBS.
Substrate solution was then added to all wells, the plate
was placed on mixer for 1 min, and a pre-read was
collected. Human factor Xa (Enzyme Research Labs
HFXa1240, titrated) was thawed at ambient temperature
immediately before use and aliquoted into vehicle at
0.4 U/ml. The factor Xa solution was then aliquoted
(20 ll) using a multi-channel pipette to start the enzyme
reaction. Plates were read on a Spectromax 250 plate
reader (Molecular Devices, Sunnyvale, CA) at 405 nm
kinetically for 10 min at 11 s/cycle, immediately followed
by an end-point read. Slopes of OD/min were plotted
against the known concentrations of calibration standards
and samples of unknown concentration were interpolated
from four parameter logistic fits of standard curves.
Compound concentrations of ex vivo samples from dosed
animals were used in SigmaPlot for the calculation of
AUCs using a SigmaPlot rev 8.1 area algorithm.
3. (a) Jones, S. D.; Liebescheutz, J. W.; Morgan, P. J.;
Murray, C. W.; Rimmer, A. D.; Roscoe, J. M. E.;
Waszkowycz, B.; Welsh, P. M.; Wylie, W. A.; Young, S.
C.; Martin, H.; Mahler, J.; Brady, L.; Wilkinson, K.
Bioorg. Med. Chem. Lett. 2001, 11, 733; (b) Sheehan, S. M.;
Masters, J. J.; Wiley, M. R.; Young, S. C.; Liebescheutz, J.
W.; Jones, S. D.; Murray, C. W.; Franciskovich, J. B.;
Engel, D. B.; Weber, W. W.; Marimuthu, J.; Kyle, J. A.;
Smallwood, J. K.; Farmen, M. W.; Smith, G. F. Bioorg.
Med. Chem. Lett. 2003, 13, 2255.
4. Masters, J. J.; Weber, W. W.; Franciskovich, J. B.; Wiley,
M. R.; Halligan, N. R.; Camp, N. P.; Jones, S. D.;
Richards, S.; Lyons, A.; Morgan, P. J.; Liebescheutz, J. W.;
Murray, C. W.; Young, S. C.; Chirgadze, N. Y.; Briggs, S.
L.; Craft, T. J.; Towner, R. D.; Smallwood, J. K.; Smith, G.
F. Abstract of Papers, 224th National Meeting of the
American Chemical Society, Boston, MA, August 18–22,
2002; American Chemical Society: Washington, DC, 2002;
MEDI-284.
5. Smith, G. F.; Shuman, R. T.; Craft, T. J.; Gifford-
Moore, D. S.; Kurz, K. D.; Jones, N. D.; Chirgadze,
N.; Hermann, R. B.; Coffman, W. J.; Sandusky, G. E.;
Roberts, E.; Jackson, C. V. Semin. Thromb. Hemost.
1996, 22, 173.
6. CLOGP, version 4.72, Daylight Chemical Information
Systems Inc., Aliso Viejo, CA 92656.
7. Pharmacodynamic surrogate of Xa inhibitor oral exposure.
Fisher rats were fasted overnight, followed by a single
administration of test compounds suspended in acacia or
CMC/Tween 80, by oral gavage. Blood was collected at
specified time-points in citrate for generation of plasma at
2000g for 15 min at 10 ꢁC. Compounds were weighed out
to the nearest 0.001 mg, adjusted for salt form and